Biochemistry
December 26, 1978
Volume 17, Number 26
0 Copyright 1978 by the American Chemical Society
Kinetic Mechanism of Escherichia coli Carbamoyl-Phosphate
Synthetase?
Frank M. Raushel, Paul M. Anderson, and Joseph J. Villafranca*
ABSTRACT: The
kinetic mechanism of Escherichia coli carbanioyl-phosphate synthetase has been determined at pH 7.5,
25 OC. With ammonia as the nitrogen source, the initial velocity and product inhibition patterns are consistent with the
ordered addition of MgATP, HC03-, and "3.
Phosphate
is then released and the second MgATP adds to the enzyme,
which is followed by the ordered release of MgADP, carbamoyl
phosphate, and MgADP. With glutamine as the ammonia
donor, the patterns are consistent with a two-site mechanism
in which glutamine binds randomly to the small molecular
weight subunit producing glutamate and ammonia. Glutamate
is released and the ammonia is transferred to the larger subunit. Carbamoyl-phosphate synthetase has also been shown
to require a free divalent cation for full activity.
Carbamoyl-phosphate synthetase from Escherichia coli
catalyzes the following reaction:
increases the number of possible kinetic schemes. Since there
are two molecules of MgATP used in the reaction sequence,
it is also of interest to determine if there are two separate
MgATP sites and if these sites can be distinguished kinetically.
Previous kinetic studies with carbamoyl-phosphate synthetase have been limited to the enzyme from frog and bovine
liver (Fahien & Cohen, 1964; Guthorlein & Knappe, 1969;
Kerson & Appel, 1968; Elliot & Tipton, 1974b,c). The liver
enzyme uses only ammonia as the nitrogen source and is activated by N-acetyl-L-glutamate. In the most complete kinetic
study, Elliot & Tipton (1 974b,c) have proposed an ordered
quad-quad mechanism for bovine iver carbamoyl-phosphate
synthetase based on initial velocity and product inhibition experiments. A different kinetic mechanism for the glutaminedependent carbamoyl-phosphate synthetase from E . coli is
proposed in this report. The mechanism is also based on initial
velocity and product inhibition experiments. The proposed
mechanism is consistent with all of the kinetic data and the
partial reactions catalyzed by carbamoyl-phosphate synthetase
from E. coli.
2MgATP
+ HCO3- t L-Gln
-
2MgADP
t carbamoyl-P P,
+ + L - G ~(1)
Ammonia can also be used as the nitrogen source. Anderson
& Meister (1965) have proposed that the enzyme catalyzes
the overall reaction through a series of three partial reactions:
MgATP t H C 0 3 -
-+
-
MgADP t -OCOOP032-
-OCOOP032- t NH3
NH2COO- t MgATP
+
NH2COO-
+ Pi
(2)
(3)
NH2COOP032- t MgADP (4)
Carbamoyl-phosphate synthetase has a molecular weight of
approximately 180 000 and is a dimer of two nonidentical
subunits (Matthews & Anderson, 1972; Trotta et al., 1971).
The smaller subunit of molecular weight 48 000 contains the
binding site for glutamine. The larger subunit of molecular
weight 130 000 contains the binding sites for the rest of the
substrates and allosteric modifiers (Trotta et al., 1971).
Carbamoyl-phosphate synthetase is a challenging enzyme
to study by kinetic techniques because the large number of
substrates and products in the overall reaction significantly
t From the Department of Chemistry, The Pennsylvania State University, University Park, Pennsylvania 16802, and the Department of
Biochemistry, School of Medicine, University of Minnesota-Duluth,
Duluth, Minnesota 55812. Received August 3, 1978. Supported in part
by Grants from the Public Health Service, GM22434 (P.M.A.) and
AM21785 (J.J.V.), and the National Science Foundation, PCM78-07845
(J.J.V.). This work was done during the tenure of an Established Investigatorship from the American Heart Association awarded to J.J.V.
0006-2960/78/0417-5587$01 .OO/O
i
Materials and Methods
Carbamoyl-phosphate synthetase was isolated from E . coli
according to the method of Matthews & Anderson (1972).
Ornithine transcarbamoylase was a gift from Dr. Margaret
Marshall. All other compounds were obtained from Sigma.
Enzyme Assays. Carbamoyl-phosphate synthetase activity
was measured spectrophotometrically using a pyruvate kinase-lactate dehydrogenase coupling system. A Beckman DU
monochrometer equipped with a Gilford 2000 optical density
converter and a 10-mV recorder was used to follow the reaction
at 340 nm.
0 1978 American Chemical Society
5587
5588
B I O C H E M IS T K Y
I : Variation of Mgzt and Mn2+ at a concentration of ATP of
10 mM. Conditions: 100 mM KCI, 50 m M Hepes, pH 7.5, 10 mM
HC03-, 10 mM glutamine, 25 OC. The velocities are in arbitrary
FIGURE
units.
For activity measurements in the absence of added MgADP,
each 1.0-mL cuvette contained 50 mM Hepes,' p H 7.5,33 pg
each of salt-free lactate dehydrogenase and pyruvate kinase,
I m M phosphoenolpyruvate, 0.2 m M NADH, 100 m M KCI,
10 mM ornithine, 15-20 mM excess MgC12, and various
amounts of the different substratea and inhibitors. When high
concentrations of NH4+ or phosphate were used, the ionic
strength was kept constant with KCI. Assays were conducted
a t 25 "C arid the carbamoyl-phosphate synthetase (1-10 p g )
was added last to initiate the reaction.
In the presence of added MgADP, activity was measured
by coupling the production of carbamoyl phosphate with ornithine transcarbamoylase. Each 1 .O-mL reaction vial contained 50 mM Hepes, pH 7 . 5 , 10 mM ornithine, 15 m M excess
MgC12, 100 mM KCI, 10 units of ornithine transcarbamoylase,
and various amounts of substrates and inhibitors. At equal time
intervals, aliquots of the reaction mixture were quenched with
0.25 mL of 10% trichloroacetic acid. The citrulline concentration was measured colorimetrically at 460 nm (Ceriotti,
1974).
Data A n a l y w . The kinetic data were analyzed using the
Fortran programs of Cleland (1967). Initial velocity data
conforming to an intersecting pattern were fitted to eq 5 and
a parallel pattern to eq 6. Competitive, uncompetitive, and
noncompetitive inhibition patterns were fitted to eq 7 , 8, and
9, respectively. When the type of pattern was in doubt (parallel
or intersecting), both possible fits were tried and a comparison
of the CT values (square root of average residual least square)
was made to determine the best fit. In no case did a set of data
fit both patterns equally well. The nomenclature used in this
paper is that of Cleland (1 963).
,I
=
VAB
~
K,,Kb
c=
c=
+ KbA + K,B + AB
VAS
KbA
+ K,B + AB
v'4
K(l
+ (I/KIb))+A
(5)
(6)
(7)
I Abbreviations used: Hepes, A;-2-hydroxyethylpiperazine-N'-2ethanesulfonic acid; EPR. electron paramagnetic resonance, PEP, phosphoenolpyruvate.
' I r M q A T P l , mM-'
Variation of MgATP at a free hlgZi conccnlration of 15 ntM.
Conditions: 10 mMglutamine, SO mM Hepes, pH 7.5, 100 nib1 KCI. 25
O C , 2.5 or 25.0 mM HCOj-. The velocities are i n arbitrar) units.
FIGURE 2:
Results
Diualent Curion Requirements. Shown in Figure 1 is the
effect on the reaction rate when the concentration of Mg2+or
Mn2+ is varied a t an ATP concentration of 10 mM. With
Mg2+ it is clear that carbamoyl-phosphate synthetase requires
more Mg2+ for full activity than required to complex all of lhe
nucleotide. With Mn", however, the optimal level of Mn2+
is about equivalent with the ATP concentration and additional
Mn2+ is inhibitory. These data suggest the presence of an additional metal ion binding site other than the metal ATP
site(s). Thus, all kinetic experiments were conducted with
excess Mg2+ present i n order to complex the nucleotides and
to also produce the additional enzymatic activatioii.
Variation of MgATP. The variation of activity with
MgATP concentration at 15 niM excess Mg2+ is shown in
Figure 2 at saturating and lionsaturating levels of HCO3--.As
can be seen from the data, the double-reciprocal plot is linear
from 10 to 300 p M MgATP.
Initial Velocity Patterns. To establish the kinetic mechanism
of E. coli carbamoyl-phosphate synthetase, the various kinetic
patterns were determined for every possible substrate pair.
Glutamine vs. either HCO3'- or MgATP gives parallel initial
velocity patterns. intersecting patterns are obtaiiied with
ammonia vs. zither HCOj-' or MgATP. However, if the
HCO3- level is raised from 1 mM to 30 m M , the N H 4 + vs.
MgATP pattern changes to one that is parallel. The MgATP
vs. HCO3- pattern is intersecting. The kinetic constants from
fits of the data to eq 5 or 6 are listed in Table I . There Is good
agreement for the K,, values of a given substrate in the various
experiments.
Product Inhibition Patterns. Product inhibition studies were
also carried out in order to establish the kinetic mechanism.
Inorganic phosphate is competitive vs. MgATP and noncompetitive vs. HC03- at low glutamine. At saturating glutamine,
phosphate is an uncompetitive inhibitor vs. HCO3-. Carbamoyl phosphate is uncompetitive vs. MgATP, HC03--,and
NH4+. MgADP is a h e a r noncompetitive inhibitor vs.
MgATP, HC03--, and NHd+. At saturating iIco3-'.the
MgADP vs. NH4+ pattern is ancompetitive. The inhibition
of Pi vs. NH4+ could riot be done because of the formation of
a precipitate. Glutamate did not significantly inhibit the reaction at concentrations up to 100 mM. The kinetic constants
from fits of the data to eq 7, 8, and 9 appear in Table IT.
Discussion
From the data presented in this paper, it is apparent that
carbamoyl-phosphate synthetase has a requirement for free
Mg2* in addition to MgATP for full activity under the assay
MECHANISM OF CARBAMOYL-PHOSPHATE SYNTHETASE
VOL.
1 7 , NO. 26, 1 9 7 8
5589
TABLE I: Initial Velocity Patterns.O
fixed
substrate
pattern
type
Gln, 10 m M
H C 0 3 - , 20 m M
HCO3-, 30 m M
intersecting
parallel
parallel
intersecting
parallel
intersecting
varied substrates
MgATP vs. HCO3MgATP vs. Gln
MgATP VS. NH4+
MgATP VS.NH4+
HCO3- vs. Gln
HCO3- vs. NH4+
0
HCO3-, 1 m M
MgATP, 5 m M
MgATP, IO mM
app Michaelis constants (mM)
HC03Gln
MgATP
NH4+
1.8 f 0.1
0.25 f 0.01
0.19 f 0.01
0.20 k 0.01
0.21 f 0.07
0.13 f 0.01
I80 f 10
130 f 40
0.16 f 0.1
2.0 f 0.1
2.3 f 0.5
170 f 30
pH 7.5, 25 OC, 100 mM KCI, 15 mM excess Mg*+. App, apparent.
~~
TABLE 11: Product Inhibition C0nstants.O
variable
substrate
inhibitor
MgATP
phosphate
HCO3-
phosphate
HCO3-
phosphate
MgATP
carbamoyl-P
HC03-
carbamoyl-P
NH4+
carbamoyl-P
MgATP
MgADP
HCO3-
MgADP
NH4+
MgADP
NH4+
MgADP
fixed
substrate (mM)
inhibition
HCO3-, 15
glutamine, 10.0
MgATP, 0.05
glutamine, 0.10
MgATP, 0. IO
glutamine, 25.0
HCO3-, 1.0
glutamine, 1.O
MgATP, 2.0
glutamine, 1.O
MgATP, 2.0
HCO3-, 10.0
HCO3-, 10.0
glutamine, 10.0
MgATP, 1.0
glutamine, 10.0
MgATP, 1.33
HCO3-, 1.0
H C 0 3 - , 50.0
MgATP, 1.33
Kii (mM)
Kis (mM)
52 f 3
C
67 f I O
NC
uc
uc
uc
uc
7 7 f 17
129 f 6
12.2 f 0.6
11.1 f 0 . 5
13fl
NC
2.1 f 0.1
0.15 f 0.01
NC
0.9 f 0.1
1 .o f 0.2
NC
2.6 f 0.1
11 f 2
uc
0.99 f 0.07
pH 7.5, 100 m M KCI, 25 "C, Kis, slope innibition constant; Kii, intercept inhibition constant; C, competitive inhibition; NC, noncompetitive
inhibition; UC, uncompetitive inhibition.
(I
conditions described. At 10 mM ATP, at least 5 mM excess
Mg2+ is required for maximal activity (Figure 1). With Mn2+
as the divalent metal, maximal activity is attained at a concentration of Mn2+ approximately equivalent to the ATP
concentration. At this concentration of Mn2+, the proposed
free divalent metal ion site and the ATP must be nearly saturated with metal. Therefore, the binding constant at the free
divalent metal ion site must be tighter with Mn2+ than with
Mg2+ because Mn2+and Mg2+ have about equal affinities for
ATP (Sillen & Martell, 1971). In preliminary binding studies
with Mn2+ using EPR techniques, we find that carbamoylphosphate synthetase binds 1 mol of Mn2+ with a dissociation
constant of approximately 40 pM, and there are also a number
of other much weaker sites.* The tight site is assumed to be the
divalent cation site needed for activity, and the weaker sites
are those that cause inhibition at higher concentrations of
Mn2+.The specific function of this free metal ion site has not
been determined, nor has the subunit to which it binds. The
presence of a free divalent metal ion binding site in addition
to a MgATP binding site has also been demonstrated for
pyruvate kinase (Gupta et al., 1976), glutamine synthetase
(Hunt et al., 1975), and PEP-carboxy kinase (Foster et al.,
1967). A requirement for excess Mg2+ for maximum activity
of carbamoyl-phosphate synthetase from other sources has
Raushel, F. M., Rawding, C . J., Anderson, P. M., & Villafranca, J.
J . (unpublished experiments).
been reported (Elliot & Tipton, 1974b; Kerson & Appel, 1968)
and similar preliminary observations have been noted with the
E . coli enzyme (Anderson, 1977; Powers & Meister, 1978).
When the excess Mg2+ concentration is maintained at 15
mM, the double-reciprocal plot for ATP is linear down to at
least 10 pM either at saturating or nonsaturating levels of
HC03-. Earlier reports of nonlinear reciprocal plots (Anderson & Meister, 1966) are probably due to inadequate Mg2+
levels to saturate both the enzyme and ATP. Thus, inhibition
by free ATP may have produced the observed nonlinearity.
Since two molecules of ATP are used in the overall reaction,
nonlinear reciprocal plots could be expected because of squared
terms for ATP in the rate equation unless there is an irreversible step (either product release or saturation of an intervening substrate) between the addition of the two molecules
of ATP. Since linear reciprocal plots are obtained, the addition
of the two molecules of ATP must therefore be separated by
a product release step and/or the addition of "3.
Kinetic Mechanism. The following kinetic mechanism is
consistent with all of the initial velocity and product inhibition
data.
Mg:TP
H Y 3
Nr3
P, M g A T P
t
Gln
Glu
1
MgADP
carbamojlP
t
t
",i"'
..S90
BI
oc H E M I STR Y
RAUSHEL, ANDERSON, A N D VILLAFRANCA
(12)
experiment) that shows that the ATPase and carbamoyl
phosphate synthesis reactions can be inhibited under certain
conditions with Ellman's reagent, while the ATP synthesis
reaction is increased almost twofold.
A recent pulse-chase type experiment has been reprted by
Powers & Meister (1978) which indicates that at high ATP
concentrations (14 mM) it is possible to populate both ATP
sites before the addition of ammonia. However, in those experiments it cannot be determined whether PI has left the enzyme surface before the addition of the second ATP (with
bound C02 as the activated form of HCO3-). It has been
suggested by Sauers et al. (1975) that bound C02, rather than
carboxy phosphate, is the intermediate that reacts with the
ammonia to form carbamate.
An alternative explanation is that the mechanism is partially
random at the point of addition of the two molecules of ATP.
However, the pathway in which both ATP molecules bind
before NH3 cannot be predominant because there is no evidence for nonlinearity in the MgATP double reciprocal
plots.
It should also be mentioned that the pulse chase experiments
of Powers and Meister were conducted in phosphate buffer and
at low Mg2+ levels, while the experiments reported here were
done in Hepes buffer, 15 mM excess Mg2+, and 10 mM ornithine. All of these reagents are known to affect the binding of
ATP to carbamoyl-phosphate synthetase. It is therefore possible for each pathway to predominate under different reaction
conditions.
The proposed kinetic mechanism in this paper is different
from the one suggested by Elliot & Tipton (1974b,c). They
have proposed the ordered addition of ATP, HC03-, ATP, and
NH3 followed by the ordered release of carbamoyl phosphate,
ADP, ADP, and PI for the beef liver enzyme. Their mechanism
is based in part on a nonlinear reciprocal plot for MgATP at
low bicarbonate and nonlinear inhibitions by MgADP. These
results are not seen with E. coli enzyme. Since their results
were obtained with a free Mg2+ concentration of 1 mM, the
observed nonlinear plots could be due to inhibition effects by
uncomplexed nucleotides and/or nonsaturation of the free
metal ion site. In addition, their mechanism does not account
for the synthesis of ATP from ADP and carbamoyl phosphate
in the absence of P, since PI is proposed to be the last product
to leave the enzyme in the forward reactibn. (It must therefore
be the first to add in the reverse reaction.) However, these
enzymes are from different sources and they have been shown
to be different in many respects. Therefore, although they
catalyze the same reaction, it is quite possible that the liver
enzyme and the E. coli enzyme have different kinetic mechanisms.
In summary E . coli carbamoyl-phosphate synthetase has
been shown to require free divalent metal ions for full activity.
The initial velocity and product inhibition patterns are consistent with the ordered addition of MgATP, HC03-, and
NH3 followed by the release of PI. The second molecule of
MgATP adds to the enzyme which is followed by the ordered
release of MgADP, carbamoyl phosphate, and MgADP. With
glutamine as the nitrogen source, the data are consistent with
a two-site mechanism in which glutamine binds randomly to
the small subunit, glutamate is released, and NH3 is transferred to the large subunit after the reaction of MgATP and
HC03-.
The strong inhibition of carbamoyl-phosphate synthetase by
Jiadenosine pentaphosphate (APsA) (Powers et al., 1977)
:;uggests two ATP sites and this is confirmed by the above kirietic mechanism. Also consistent with the existence of two
.ATP sites is an experiment (P. M. Anderson, unpublished
References
Anderson, P. M. (1977) Biochemistry 16, 583.
Anderson, P. M., & Meister, A. (1965) Biochemistry 4,
2803.
By using the rules formulated by Cleland (1963), the rea,,,::ning used to derive the above mechanism is as follows:
I . The initial velocity patterns for bicarbonate vs. both
S lgATP and NH4+ are intersecting and the MgATP vs. NH4+
?;.ittern is intersecting at low HC03- but parallel at saturatihp
;iCO3-. These results demonstrate that HCO3- binds to the
,nzyme between the addition of MgATP and NH4+.
2. The MgATP double-reciprocal plot is linear at both sat::[-ating and nonsaturating levels of HC03-, indicating that
.: product release step must separate the addition of the two
aolecules of MgATP.
3. Phosphate is a competitive inhibitor vs. MgATP and thus
;hey bind to the same enzyme form.
4. Phosphate is a noncompetitiveinhibitor vs. HC03- at low
irlutamine and uncompetitive at saturating glutamine indicating that the ammonia derived from glutamine binds be:ween the addition of HCO3- and the release of Pi.
5 . Carbamoyl phosphate is an uncompetitive inhibitor vs.
, ~ l three
l
substrates indicating that there is a product released
immediately before and after the release of carbamoyl phos:ih a t e.
6. The inhibition patterns for MgADP are linear noncom::etitive vs. MgATP, HC03-, and NH4+, but the NH4+ vs.
?fgADP inhibition pattern is uncompetitive if the HC03- level
i s raised to 50 mM from 1 mM. These results show that the
oinding of HC03- precedes the binding of ammonia. Since all
:if the inhibition patterns are linear, the release from the eniyme of the two molecules of MgADP must be separated by
release of another product (in this case carbamoyl phos!hate).
7. The glutamine vs. MgATP and HCO3- initial velocity
>dotsare both parallel. This is different from the result when
,. mmonia is used. The reason for the parallel patterns is the fact
:hat the binding site for glutamine is on a different subunit than
1 he rest of the reaction and the ammonia must therefore be
iiansferred from one subunit to the other. The parallel pattern
m u l t s from the release of glutamate before the NH3 reacts
t the other subunit. This situation is similar to the biotin eni-ymes which have a two site ping-pong mechanism (Northrop,
i969; McClure.et al., 1971). In this case ammonia is being
:!ansferred instead of carboxy biotin. The results with gluta?kine are consistent with the initial velocity data using am:.'ioniaas a substrate if it is proposed that the ammonia is not
*:,ansferreduntil ATP and HCO3- have reacted on the other
,*:ibunit.The rate equation for the proposed kinetic mechanism
:.;>ingNH3 as the nitrogen source has been derived using the
iiethod of King & Altman (1956). The predicted initial ve:;city and product inhibition patterns from the derived equai kin agree with the experimental patterns. The initial velocity
tiid product inhibition patterns for this particular mechanism
have also been derived by Elliot & Tipton (1974a).
'The proposed kinetic mechanism is consistent with the three
1)artialreactions catalyzed by carbamoyl-phosphate synthetase
{Anderson & Meister, 1966).
~
€KO,-
MgATP
carbamoyl-P
+ MgADP
glutamine
-
-
+ Pi
(10)
MgATP + HCO3- + NH3
MgADP
(11)
glutamate
+ NH3
METHIONYL-tRNA SYNTHETASE
Anderson, P. M., & Meister, A. (1966) Biochemistry 5,
3 157.
Ceriotti, G. (1974) Methods of Enzymatic Analysis (Bergmeyer H., Ed.) p 691, Academic Press, New York.
Cleland, W. W. (1963) Biochim. Biophys. Acta 67, 104.
Cleland, W. W. (1967) Ado. Enzymol. 29, 1.
Elliot, K. R.F., & Tipton, K. F. (1974a) Biochem. J . 141,
789.
Elliot, K. R. F., & Tipton, K. F. (1974b) Biochem. J . 141,
807.
Elliot, K. R. F., & Tipton, K. F. ( 1 9 7 4 ~ )Biochem. J . 141,
817.
Fahien, L. A., & Cohen, P. 0. (1964) J. Biol. Chem. 239,
1925.
Foster, D. O., Lardy, H. A., Ray, P. D., & Johnston, J. B.
( 1 967) Biochemistry 6, 21 20.
Gupta, R. K., Fung, C. H., & Mildvan, A. S. (1976) J . Biol.
Chem. 251; 2421.
Guthohrelein, G., & Knappe, J. (1969) Eur. J. Biochem. 8,
207.
VOL.
17,
NO.
26, 1978
5591
Hunt, J. B., Smyrniotis, P. Z., Ginzburg, A., & Stadtman, E.
R. (1975) Arch. Biochem. Biophys. 166, 102.
Kerson, L. A., & Appel, S. H. (1968) J. Biol. Chem. 243,
4279.
King, E. L., & Altman, C. (1956) Phys. Chem. 60, 1375.
Matthews, S. L., & Anderson, P. M. (1972) Biochemistry 1 1 ,
1176.
McClure, W. R., Lardy, H. A., Wagner, M., & Cleland, W.
W. (1971) J . Biol. Chem. 246, 3579.
Northrop, D. B. (1969) J . Biol. Chem. 244, 5808.
Powers, S. G., & Meister, A. (1978) J. Biol. Chem. 253,
800.
Powers, S. G., Griffith, 0.W., & Mesiter, A. (1977) J . Biol.
Chem. 252, 3558.
Sauers, C. K., Jencks, W. P., & Groh, S. (1975) J . Am. Chem.
SOC.97, 5546.
Sillen, L. G., & Martell, A. E., Ed. (197 1) Chem. SOC.,Spec.
Publ. No. 25, Suppl. 1 , 650.
Trotta, P. P., Burt, M. E., Haschemeyer, R. H., & Meister,
A. (1971) Proc. Natl. Acad. Sci. U.S.A. 68, 2599.
Mechanism of Aminoacylation of Transfer RNA. A Pre-Steady-State
Analysis of the Reaction Pathway Catalyzed by the
Met hionyl-t RNA Synthetase of Bacillus stearot hermophilust
Roderick S. Mulvey* and Alan R. Fersht*
ABSTRACT:The rate of formation of the product MettRNAMet,catalyzed by the methionyl-tRNA synthetase from
Bacillus stearothermophilus, has been determined in both the
pre-steady-state and steady-state phases by rapid sampling
techniques. A minimum estimate of the formation of the intermediate Met-AMP, in the presence of tRNAMet,has been
obtained from the kinetics of ATP-pyrophosphate exchange.
The individual rate constants for the formation and interconversion of enzyme-bound intermediates have been directly
measured by stopped-flow fluorescence. The two approaches
have been combined to give a description of the reaction
pathway. The pre-steady-state rate of methionyl transfer from
E.(Met-AMP)2 to tRNA was measured over a range of pH
and with two species of tRNAMet.Under all conditions, this
rate constant was identical to the initial rate of charging of
tRNAMetin the steady-state phase (for example: 2.3 s-' at pH
7.78 with tRNAMetM).Therefore, methionyl transfer, or a
conformational change immediately preceding transfer, is the
rate-determining step. The rate of formation of methionyl
adenylate in the presence of saturating concentrations of all
substrates and the rate of dissociation of tRNAMetfrom its
complex with E.(Met-AMP), are both faster than the proposed
rate-determining step (25 and 12 S - I , respectively, at pH 7.78
with tRNAMetM).The k,,, for ATP-pyrophosphate exchange
under the same conditions, 1 1 s-I, is significantly less than the
rate constant for the formation methionyl adenylate measured
by stopped-flow fluorescence (25 s-l).
T h e aminoacyl adenylate mechanism has been established
beyond reasonable doubt for several enzymes (Fersht and
Jakes, 1975; Fersht and Kaethner, 1976; Fasiolo and Fersht,
1978; Gangloff and Fersht, 1978; Mulvey et al., 1978). In this
paper we have chosen one enzyme, the methionyl-tRNA synthetase from Bacillus stearothermophilus. for a detailed kinetic investigation. This enzyme, a dimer (2 X 82 000), possesses an editing mechanism which reduces the incidence of
mischarging of tRNAMetwith noncognate amino acids such
as homocysteine and a-aminobutyrate (Fersht and Dingwall,
unpublished results). In addition, it is particularly suitable for
study by pre-steady-state kinetics, since, like the equivalent
enzyme from Esrherichia coli (Hyafil et al., 1976), its intrinsic
fluorescence shows great sensitivity to events at the active site.
A combination of stopped-flow fluorescence and quenchedflow experiments has made possible the identification and
study of the following steps in the reaction pathway: methionyl
adenylate formation, tRNA binding, and aminoacyl transfer
from the methionyl adenylate complex to tRNA. Since the
formation of methionyl adenylate in the presence of tRNA can
be followed by stopped-flow fluorescence;the rate constant for
this step can be measured directly instead of inferring its value
from the kinetics of pyrophosphate exchange. Comparison of
the rates of these individual steps with the overall rate of aminoacylation of tRNA has established that methionyl adenylate
does accumulate during the turnover of substrates and that
aminoacyl transfer from this complex to tRNA is the ratelimiting step.
From the MRC Laboratory of Molecular Biology, Cambridge CB2
2QH, England. Received May 12, 1978: revised manuscript received
August 8. 1978.
0006-2960/78/0417-5591$01 .OO/O
0 1978 American Chemical Society